We believe that the AI agents, it will reach those parts of the community that don't want to pick up a phone. I was quite surprised our nurse hotline is often described by people as the best service that they'd never heard of. The magic sauce is really good information on a trusted platform, can provide great outputs for people. Accurate, evidence-based, but also the public ultimately can rely on.
SPEAKER_02Hi, and welcome back to Missions to Movements. I am your host, Dana Snyder, and we are coming to you live from the studio at the Microsoft Global Nonprofit Leaders Summit. And I have with me a very special guest, Mike Osborne, Chief Information Officer at the Cancer Council Victoria. Mike, thank you for joining me.
SPEAKER_00Thanks for having me.
SPEAKER_02Of course. Before we dive into the amazing work that you do, I would love to help you have you share a little bit about who you are, what your role is, what does a day-to-day look like if one exists, and a little bit about what just genuinely lights you up about this work you get to do.
SPEAKER_00Maybe I'll start with Cancer Council Victoria. We are a charity based out of Melbourne, Australia. I've been part of a federation of uh cancer councils in in the country. We focus on cancer prevention, early detection, and then uh cancer information and support for those affected by cancer. And finally, we uh focus on research, uh, whether that's funding external cancer research or doing our own epidemiological research. But of course, all that's supported by our amazing fundraisers who are aiming to give every Victorian an opportunity to contribute to uh our vision of a cancer-free future.
SPEAKER_02Love that.
SPEAKER_00My role as the CIO really is two parts. There's the day-to-day bit of just making sure that all of our great people have the tools and
the technologies to safely do what they do. And then probably more excitingly, my team leads our digital transformation, trying to make sure we've got the best possible capabilities that really empower and amplify the impact of the amazing people who work in the organization.
SPEAKER_02Amazing. If there was to say a favorite aspect, what do you love to focus on the most or gives you excitement?
SPEAKER_00For me, the the best part about the job is when I get to work directly with the folks who are making a difference out in the community, whether that's our cancer nurses who are helping people through a very difficult time in their lives, the researchers finding out incredible insights into cancer. It's where we get to engage directly with the people who do such amazing work helping folks in the community.
SPEAKER_02Amazing. And you've recently shared, I love LinkedIn. So I was reading a LinkedIn article about this, and you described your team keeping a vision board of what's possible. A, I think everyone should do that. But you talk about from prevention to early detection to ongoing support. When did you really start exploring AI as a tool, as an infrastructure that would help you for specifically you had this supportive care hotline? When was that coming into existence? And why did the helpline really become one of the first places of focus?
SPEAKER_00Yeah, I think like many probably, when ChatGPT came out, what, about three years ago now, a short play of that, and you realize the potential amplification of our mission that could be achieved through it. So we spent a fair amount of time thinking about and envisioning where can this technology can help us achieve our mission. And we know that one in two people will get impacted by cancer in their lives. And so there's an immense need for support out there. We could never have enough cancer nurses to provide the one-on-one support to everybody. So, how can we use the AI tools to amplify, become their co-pilots to help them do their job, but then ultimately
make those public-facing and provide trustworthy information in a really accessible format for people? So that was our first big idea.
SPEAKER_02Yeah.
SPEAKER_00We started working on that and created some pilots, but then pivoted largely due to funding sources. But when the Victorian government, the consumer Department of Consumer Affairs, gave us a very generous grant to work in the financial counseling space, we pivoted to how can we empower our financial counselors who help people work through what can be a very difficult financial situation when diagnosed with cancer and help people navigate that better.
SPEAKER_02Right. That's incredible. This has all been in the past three years. Fascinating. How has the team shifted and or the different tools now been supportive in making this possible?
SPEAKER_00So one of the things there's a number of things we worry about. It's so fundamentally important that the information we share is fact-based, high quality. And so we settled on using Microsoft and the uh Azure AI services for this because we know that there are controls built in for us. And so we believe the magic sauce is really good information on a trusted platform can provide great outputs for people. And so that's where we're focused a lot on is how do we make sure that the information we provide is accurate, evidence-based, and something both our counselors but also the public ultimately can rely on.
SPEAKER_02That's right. How have you seen a transformation happen since putting this into place?
SPEAKER_00Well, we're still in in pretty early days. So the financial counseling assistant is in use with our financial our financial counselors right now in sort of a testing and trying mode. Okay. But what we're seeing is it's helping them find information quicker. It's helping them produce uh follow-up documentation and emails quicker and is performing very accurately.
So we really hope, and we're moving into a a trial, a public trial shortly, and then aiming to launch publicly at the end of the year.
SPEAKER_02Wow.
SPEAKER_00Where we really think, you know, we've only got a handful of financial counselors, and there's so many people who need help that can we offload some of the more straightforward requests and needs to the AI agent and to let our counselors then focus on the more complex situations that that some people are have.
SPEAKER_02That's fabulous. One, congratulations. That's a big feat, and I'm sure it takes a lot of intricacies within the organization to make possible. So the helpline experience before in this case, talking about the counselors, like something that can never be manufactured is is time and presence, right? The human interaction point. So for the counselors themselves, what have they said about the experience? And I mean, this is f fantastic that you're gonna be able to hopefully scale in such a way, but have they said in their words, maybe what the experience has been like?
SPEAKER_00Yeah, and I think what we're seeing is the counselors on a day-to-day basis can have requests from a very simple how do I apply for something through to very complex cases where with through the cancer treatment, individual may not be able to work, they may be having trouble paying their rent, putting food on the table. And so there's a a need for our counselors to have a really deep discussion about the circumstances of that individual, what their insurances might be, what government supports they might qualify for, etc. And so what the AI is letting them do is quickly find some of that information. You know, in the in in pre-AI, they would have been jumping to different websites and looking for processes that you know the that the individual may need to go through with the government. They may be researching what insurances are included in in their uh retirement. So it takes a long time trying to search and source information, which distracts them from the really important conversation that they're having with that individual who's in a who's having a really tough go. So if we can make that easier for the counselors and have the AI agent source the information for them, they can spend their time worrying about the individual that's on the other line end of that phone line.
SPEAKER_02Wonderful. I want to just listener, for those of you or viewing, is to think about such a great case study into what's an actual problem or scalable opportunity at your organization that technology can solve so that the people can focus on the more human connections, interactions that you need to have. Thinking about it always starts with I love today during one of the keynotes they were talking about, don't actually start with AI. Think about a problem. Think about an opportunity, and then think about how the tech can be utilized to help you solve that. One of the most unexpected findings from one of the programs that I want to bring up was an example from a youth smoking prevention program. So you realize that teenagers were more willing to open up to an AI chat bot than call an adult. Can you share this story with us a little bit?
SPEAKER_00You know, Australia's done a fantastic job of driving down smoking rates over the past several decades. And one of our key services is called Quit. It's the smoking cessation or or now tobacco cessation service for much of the country. But really troublingly, we've seen that you know the curve of reduction of tobacco use has slowed because of uh vaping and the impact that's having, and it particularly on youth. So we began a piece of work to look at how what kind of interventions we can make with youth to help them get off the vapes. And through our co-design with a youth population, I was quite surprised, because this was probably a couple of years ago now, they were much more comfortable opening up to an AI agent than picking up the phone and calling what they perceived to be a judgmental older adult. And so they really liked the idea of being able to have an open conversation with an AI
agent to help them work through it the challenges they're having.
SPEAKER_02How has that informed how you then design a chat bot in an experience like that?
SPEAKER_00So we're working really hard on tone. How do we ensure that the tone of the chatbot is supportive, speaks to them in a more youthful sort of language? Because what they don't want is, you know, a 50-year-old guy on the other end of the phone giving them a lecture about uh their vaping habits. So there's considerable work going into how we do that. The other key one is we're building it as a messaging app so that the youth can you know jump in and directly message the AI agent.
SPEAKER_02Meeting them where they are.
SPEAKER_00Right, exactly. It's an interesting journey. Uh we're now getting to the stage where we're building that infrastructure out. It's all grounded in sort of the fact behavioral science-based cessation strategies, but introduces that ability to have the conversation.
SPEAKER_02Yeah. Fascinating. We've kind of framed this conversation about being able to widen the care offering, right? Reaching people who might not otherwise get support and or may not feel comfortable in the support. Before, who do you think might have been falling through the cracks? Or imagine, I mean, the teens is a great example. Maybe they wouldn't have picked up the phone in the first place. Other areas in the organization where you've stood up now these opportunities through technology to be able to be more accessible?
SPEAKER_00You know, our traditional telephone support line is overwhelmingly used by a certain demographic, and that tends to be middle-aged women who are most likely to pick up that phone and call. We know others have different ways they want to access information. So our websites and the cancer information we provide on that is a huge resource. Interestingly, we're very concerned of the impact of the AI summaries that are appearing in all of our favorite search engines, because those are doing two things. One, they're drastically reducing the traffic coming to the site as people get their information from the summary. Yeah. But interestingly, our colleagues in Europe did a short study around the quality of those responses and found that more than half of the of the responses they saw actually had inaccuracies in them, which is very troubling.
SPEAKER_02Yes.
SPEAKER_00If someone is researching, you know, the cancer that they that a loved one might have and and treatment
and et cetera. We believe that the AI agents, if we can build a trusted AI agent, it will reach those parts of the community that don't want to pick up a phone. We can tailor them to communities that are traditionally underserved. So if if English is not your first language and you're trying to sort out a whole bunch of dense scientific text around your disease, how how can we simplify that down? How can we potentially do that in your language?
SPEAKER_02Usually the most easy to understand and usually very dense. Yeah.
SPEAKER_00If you live regionally more in a remote area, you've got unique um challenges seeking treatment. So how can we help connect you, those individuals better? So there's just so many areas that as we either directly through the AI or by freeing up the time of our people, allow us to then focus on uh specific sub-populations that really need our support.
SPEAKER_02Yeah. Are you seeing in analytical tools that more traffic is coming to your site through AI, though? Have you noticed that trend yet?
SPEAKER_00No. Like uh Like GPTs and yeah. The we we are getting some traffic from those, but overarchingly we've seen a significant decline in the overall traffic on the sites.
SPEAKER_02Aaron Powell This is a really good thing to note for organizations. When you have these great tools now accessible that are of quality and giving great information, how do you make sure people can find it?
SPEAKER_00Which is one of the big challenges in even our traditional services. Our nurse hotline is often described by people as the best service that they'd never heard of. So again, reaching people across the different channels, whether and increasingly social media to spread the message and have them connect to us. We can no longer rely on SEO as you know, sort of the primary driver for us. We have to reach out across social media as continue to work to make sure that our high-quality information is what the AI engines are referencing.
SPEAKER_02Right.
SPEAKER_00But it's an area of concern and and we're still trying to figure our way through it.
SPEAKER_02To figure it out. Yeah, I think we're all in that place. It's all relatively very new. For a CIO at a 90-year-old organization, which is incredible. How did you earn the internal trust for AI adoption, which I would imagine would be its own challenge maybe at times, or maybe not? So that's the first question. And then for all the listeners and viewers that are here, what would you say to those that are going through maybe that struggle internally?
SPEAKER_00You know, we're on a journey as everyone else is with this technology. We've taken a multi-pronged approach. We've enabled Copilot Chat for the organization as a default tool that is protected behind our information protection framework, making sure that we're not sending data out to the public engines and whatnot. So we're building trust that you can use the tool safely, but we're encouraging people just to have a go. So we're now up to about 60% of our organization are active users on Copilot Chat. That's growing every month. And so we're just trying to make it something for them that's about how they can amplify their impact. And that's sort of the base level. We're doing a lot of work with our executive team around how they can use the tools and really lead from the front. And then we're making these targeted investments in the really mission-oriented chatbots that we've been talking about here.
SPEAKER_02Yep.
SPEAKER_00And those are all being built out in the Azure AI space. So again, they're within
our tenancies. Yeah. They're grounded on our data and thus much safer to operate than some of the public stuff.
SPEAKER_02Fascinating. For an organization that might not have all the resources that you do, where would you suggest with your CIO hat on, where would you suggest that they start?
SPEAKER_00I'd go for Copilot Chat. It's a free service. And it's just thinking of things, you know, as an individual that take you time and take you away from the important stuff. And then, you know, any of us can do it. I sat down a few weeks ago. One of for my sins, one of the things I have to do is review a lot of contracts to make sure that they've got the right security provisions in them and we're not transferring data we shouldn't be, et cetera. So I sat down and spent half a day and created myself a little chatbot that I can just pump the contract into and it will review it against our standards and compare it to other good contracts and our standard templates, and it saves me, you know, hours. Hours. So just get into it, have a play.
SPEAKER_02Because I'm sure that was your favorite part of your job.
SPEAKER_00It was a lovely. I love reading 50-page legal contracts. So the more I can do less of that, yeah. While still making sure we're not we're not signing bad contracts and I can spend that time doing more exciting things.
SPEAKER_02That's right. That's right. Mike, thank you so much for your time here today and for all the work that you do. Thank you.
SPEAKER_01Thank you so much for tuning into today's episode of Missions to Movements. If you enjoyed our conversation and found it helpful, I would love for you to take a moment to leave a review wherever you're listening. Your feedback helps us reach more change makers like you and continue bringing impactful stories and strategies to the show. Don't forget to hit that subscribe button too so you'll never miss an episode. And until next time, keep turning your mission into a movement.